The job demands ‐ resources theory

The job demands -resources theory originated in the burnout literature and was first proposed by Demerouti et al. [62], and in its later development, the theory synthesized knowledge from various theories of work stress and job motivation and argued that burnout and work engagement are the result of various job characteristics [72]. The theory suggests that job characteristics can be categorized into job demands and job resources. The former relate to physical, psychological, social, or organizational aspects of work, which require sustained physical, cognitive, and emotional exertion, which in turn entails associated costs such as work overload, interpersonal conflict, and job insecurity, while the latter relate to elements of the same domains that have a motivational effect, such as coworker support, job autonomy, and performance feedback, which contribute to goal attainment and promote learning and personal growth [63, 64, 72]. Although both excessive job demands and scarce job resources can lead to burnout, reducing job demands alone will only alleviate burnout without affecting work engagement, only abundant job resources will trigger work engagement [72]. Therefore, by increasing job resources, organizations can increase employee engagement while attenuating burnout. In light of the above considerations, the current study aims to reveal the mechanism of action by which digital-HRM, as an emerging job resource in the digital era, promotes positive employee psychology and behavior while dampening negative reactions, and to further clarify how organizations can optimize the impact of digital-HRM in these areas.

Linking digital-HRM to employee innovative performance via sense of work gain

Sense of work gain emanates from the construct of the “sense of gain”, which refers to a subjective feeling that individuals based on their perceived “objective acquisition” in the process of self-improvement and continuous advancement of their situation [73]. In this context, the objective acquisition includes not only material acquisition such as salary, but also spiritual acquisition such as organizational support and personal growth. Sense of work gain entails a comprehensive evaluation of the actual rewards and value that employees perceive in their work, specifically appraising the fairness and reasonableness of the relationship between their “work effort” and “work reward” [65]. This underscores the extent of contentment derived from individual needs fulfilled through work-related benefits [66]. Prior research has demonstrated that positive job characteristics and organizational support contribute to employees’ sense of work gain, thus fostering job well-being and engendering organizational citizenship behaviors [65, 74].

Based on the job demands-resources theory, job resources are considered to stimulate employees’ work motivation processes by satisfying their basic psychological needs, fostering engagement, and ultimately enhancing creativity and performance [72, 75–78]. Within this theoretical framework, digital-HRM plays a crucial role as a job resource, significantly fostering positive psychological experiences and thereby boosting employees’ innovative performance. We posit that digital-HRM serves as a job resource in three main areas: transforming the work environment, optimizing HRM processes, and enhancing employee services [40]. Firstly, digital-HRM transforms the work environment from a physical to a digital realm, not only enhancing work flexibility and connectivity but also creating an empathetic space that supports collaboration within the organization [40, 67, 79]. This shift meets employees’ basic psychological needs for flexibility and connectivity, promoting collaboration and innovation. Secondly, digital-HRM, through the refinement and intelligent transformation of management processes, has given rise to intelligent digital service systems and automated HR activities [1], which constitute new paradigms of governance and service. This not only improves work efficiency but also satisfies employees’ needs for skill development and personalized attention by providing timely feedback and customized services [53]. Furthermore, online workflows embedded in digital-HRM increase time efficiency, allowing HR professionals to divert from tedious daily tasks to focus on enhancing employee well-being and job satisfaction [40]. This shift gives employees more opportunities to develop skills and engage in meaningful activities [1], fulfilling their needs for achievement, self-efficacy, and self-actualization. In summary, we believe that digital-HRM, as a job resource, fully meets the diversified psychological needs of employees, providing comprehensive organizational support and significantly enhancing their sense of work gain.

Furthermore, we posit that the sense of work gain brought about by digital-HRM can further enhance employee innovative performance, which refers to the extent to which employees generate and implement new and useful ideas in the workplace [80]. Based on the aforementioned analysis, the sense of work gain derived from digital-HRM is primarily manifested on three levels: the foundational level of sense of work gain, which pertains to the fulfillment of work efficiency and income; the intermediate level of sense of work gain, which pertains to the fulfillment of respect; and the highest level of sense of work gain, which pertains to the fulfillment of self-actualization [66]. These varying levels of psychological fulfillment enable employees to have more flexible arrangements of work time and location [81, 82], concentrate on meaningful and valuable work [1], receive more personalized development [81, 83], and ensure that their contributions and value are duly recognized. This clarity in self-perception and growth leads to a greater willingness to learn new skills and knowledge [84, 85], as well as to share creativity and take risks within teams and the organization, thereby leading to improved innovation performance [3, 86]. Simultaneously, research results also indicate that a direct correlation between an elevated sense of work gain and favorable workplace encounters, coupled with heightened job contentment [65]. Significantly, individuals who manifest greater job satisfaction evince a proclivity for embracing more substantial risks and showcasing heightened creativity, this predisposition culminates in the manifestation of innovative behaviors and elevated innovative performance [33, 87]. Within this context, it is posited that digital-HRM can play a pivotal role in enhancing employees’ sense of work gain, thereby engendering an upsurge in their innovative performance. As a logical consequence, the following hypothesis is formulated:

1. H1: Sense of work gain mediates the positive relationship between digital-HRM and employee innovative performance. The higher the degree of digital-HRM, the more employee innovative performance is triggered by employees based on their sense of work gain.

Linking digital-HRM to cyberloafing via relative deprivation

Relative deprivation refers to a negative subjective cognitive experience in which individuals or groups perceive themselves as disadvantaged or unfairly treated when making comparisons with other relevant groups [88]. Once individuals experience relative deprivation, it is easy to develop negative emotional as well as behavioral responses, and even trigger group behavior [89]. In terms of conceptual connotation, the meaning of relative deprivation is opposed to the meaning of sense of gain [73, 90]. Therefore, we believe that digital-HRM may diminish employees’ relative deprivation.

Similarly, according to the principles of the job demands-resources theory, job resources can alter the adverse perceptions and cognitions caused by job demands, thereby mitigating negative reactions during the evaluation process [72, 91–93]. Therefore, this study posits that digital-HRM, as an emerging job resource, can significantly change the adverse reactions caused by job demands on employees, such as feelings of relative deprivation. Firstly, digital-HRM increases the transparency of management activities through digital tools and platforms [53, 82], enabling employees to more easily access and understand decisions and policies related to their work [40]. This transparency directly alleviates the psychological stress and uncertainty caused by informational asymmetry, which are often major negative factors in traditional job demands. It helps reduce the likelihood of opaque operations or biased treatments, further eliminating concerns about improper interventions in HRM measures and alleviating feelings of relative deprivation. Meanwhile, as digital technology is progressively integrated within organizations, work processes are transforming towards higher transparency and visualization, steadily moving towards a “fully traceable” paradigm [3, 40], which grants unparalleled accessibility to information [67]. This enhances employees’ perceptions of fairness and justice, reducing their concerns about unfair treatment and the consequent feelings of relative deprivation. Secondly, digital-HRM alleviates many traditional job demands, such as repetitive tasks and time management pressures [1], through automation and intelligent technologies. Automation not only frees up time and energy for employees to engage in more creative and strategic work but also reduces the feeling of work overload [40], thereby decreasing the perception of deprivation. Overall, digital-HRM not only addresses the challenges brought by traditional job demands but also actively utilizes technological resources to optimize employees’ work experiences, thus weakening their feelings of relative deprivation.

We argue that this reduction in relative deprivation can attenuate employees’ cyberloafing, which is an IT way of idling on the job, and is often a consequence of perceived unfairness [60]. Relative deprivation, as a perception of unfair treatment, can trigger negative emotions that drive individuals to engage in inappropriate behaviors as a means to compensate for their perceived deprivation [94]. These behaviors are often aimed at quickly alleviating the sense of deprivation in the short term [95], such as addiction to online games [96] and theft [97]. Previous research has shown that cyberloafing can serve as a coping mechanism to alleviate feelings of exhaustion and emotional strain caused by workplace rejection or work-related stress [68, 69, 98].Therefore, we suggest that feelings of relative deprivation may drive employees to engage in cyberloafing as a means to cope with the emotional exhaustion. Thus, when digital-HRM reduces employees’ relative deprivation, we expect that their engagement in cyberloafing will also decrease. Combining the above analyses, we propose the following hypothesis:

1. H2: Relative deprivation mediates the negative relationship between digital-HRM and cyberloafing. The higher the degree of digital-HRM, the lower the amount of cyberloafing triggered by employees based on relative deprivation.

The moderating effect of perceived ease of technology use

The conceivable applications and advantages of digital-HRM are contingent upon the degree to which employees embrace and integrate digital technologies within their work environment [3, 99, 100]. Consequently, our investigation delved into the moderating role played by the perceived ease of technology use. This construct pertains to the extent to which employees perceive the use of a specific technology as unproblematic and devoid of complexities [101].

Drawing from the self-determination theory, it is known that when individuals’ intrinsic needs for competence, autonomy, and relatedness are met, their self-motivation and well-being are enhanced [102]. In the context of the digital age, employees’ perceived ease of technology use becomes a key factor in fulfilling these psychological needs. Firstly, the perceived ease of technology use directly strengthens the sense of autonomy in their work. When technology is perceived as user-friendly, employees can complete tasks more quickly and efficiently, reducing delays and frustrations caused by technological barriers [103]. Moreover, as the ease of using technology increases, employees feel more control and freedom of choice when using technology [104]. This enhanced autonomy encourages them to adopt and utilize technology more proactively, rather than feeling forced to adapt, thus improving job satisfaction and engagement [105–107]. Secondly, a high perception of technology’ ease of use enhances their sense of competence. When employees believe technology is easy to master and use, they are more likely to feel competent at their jobs and able to complete tasks more quickly and receive positive feedback [108, 109], thereby boosting their sense of self-efficacy and accomplishment. Indeed, using a longitudinal design, Xanthopoulou et al. [110] discovered that individuals with high self-efficacy reported higher levels of autonomy and growth opportunities over time, subsequently influencing their experiences of self-motivation and well-being. In a technology-driven work environment, a smooth technological user experience is a crucial factor in enhancing employees’ sense of competence. Lastly, the perceived ease of technology use lowers the barriers to communication and collaboration, not only enabling employees to interact more effectively with colleagues and establish and maintain positive work relationships but also making them feel more interconnected and supported [111], satisfying their social belonging and relatedness needs. It also promotes collaboration and information sharing among team members [112], contributing to improved overall team efficiency and innovative capacity. Effective participation and contribution within the team enhance employees’ sense of social value, reducing feelings of marginalization or relative deprivation.

In conclusion, the perceived ease of technology use is crucial for satisfying employees’ needs for autonomy, competence, and relatedness. The higher the employees’ perceived ease of technology use, the more likely they are to experience enhanced job satisfaction, engagement, well-being, and team support. Therefore, we believe that employees with a higher perceived ease of technology use are more likely to experience increased sense of work gain and reduced feelings of relative deprivation. Building upon the aforementioned arguments, we propose the following moderating hypotheses:

1. H3a: Perceived ease of technology use moderates the positive relationships between digital-HRM and sense of work gain, such that the relationships are stronger when perceived ease of technology use is higher rather than lower.
2. H3b: Perceived ease of technology use moderates the negative relationships between digital-HRM and relative deprivation, such that the relationships are stronger when perceived ease of technology use is higher rather than lower.

Combining H1, H2, H3a and H3b, this study further proposes two moderated mediation hypotheses. Specifically, for employees with a high-level perceived ease of technology use, they will be more inclined to embrace digital-HRM, from which they will experience more need satisfaction and thus higher sense of work gain, which will motivate them to exhibit higher innovative performance; at the same time, employees’ perception of relative deprivation will be further weakened, which will further inhibit them from committing cyberloafing behaviours.

1. H4a: The positive indirect relationship between digital-HRM and employee innovative performance through sense of work gain will be moderated by perceived ease of technology use such that the relationship will be stronger (vs. weaker) when perceived ease of technology use is higher (vs. lower).
2. H4b: The negative indirect relationship between digital-HRM and cyberloafing through relative deprivation will be moderated by perceived ease of technology use such that the relationship will be stronger (vs. weaker) when perceived ease of technology use is higher (vs. lower).

The research model of this paper is shown in Fig 1.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Theoretical framework.

https://doi.org/10.1371/journal.pone.0307195.g001

Sample and data collection

This study develops a moderated mediation model that proposes the positive effects of digital-HRM, an emerging job resource, on employee innovative performance as well as the negative effects of cyberloafing through the dual mediation of sense of work gain and relative deprivation and the moderating effects of perceived ease of technology use. To this end, we surveyed full-time employees from various industries in China (including Internet, finance, energy, construction, and retail) through the Credamo. Credamo is a large-scale intelligent online research platform which allows the distribution of questionnaires to those who meet the inclusion criteria.

Our study started on December 7, 2022 and ended on December 22, 2022. We used a questionnaire to distribute 550 questionnaires on Credamo, offering a 5¥ incentive per questionnaire. Each questionnaire included a statement informing participants of the nature, confidentiality, and objectives of the study. Since we utilized an online distribution method for the questionnaires, all participants had the right to decide whether to participate, ensuring voluntary involvement. Once participants completed the questionnaire and submitted it, they indicated their consent to participate in our study. The Academic Committee of the Research Institute of International Economics and Management, Xihua University, exempted this study from ethical review and since our study ensured the willingness of the participants to participate in the survey, the Academic Committee stated that this study did not require any written or verbal formal consent from the participants. To ensure the quality of the sample, we performed some operations before we started collecting data. Firstly, we restricted the survey population before distributing the questionnaires, and only allowed the employees of the enterprises to fill in the questionnaires; secondly, we restricted the IP addresses when distributing the questionnaires, and prohibited the same IP address from answering the questionnaires more than once; and at the same time, we also restricted the answer range, and only allowed one IP address to answer the questionnaires within 5km, so that we could avoid the situation that the group answering would interfere with the data results. Finally, we also set up interference questions to exclude the questionnaires that were not filled out carefully to further ensure the quality of the questionnaire. Moreover, to avoid common method variance [113], we conducted the survey at three points in time. At time 1, we distributed 550 questionnaires, including questions regarding digital-HRM, perceived ease of technology use, and control variables. And, 548 responses were received. At time 2, after a gap of one week, questionnaires on sense of work gain and relative deprivation were sent to the 548 initial participants. And, 523 responses were collected, generating a response rate of 95.44%. At time 3, one week after time 2, questionnaires on employee innovative performance and cyberloafing were sent to the 523 participants, 502 responses were collected. After excluding incomplete responses as well as straight-line responses and outliers, 487 valid questionnaires were obtained, with an effective response rate of 88.55%. Data are available in S1 Data.

Due to the unavailability of official data regarding the characteristics of employees in the context of digital-HRM, we refer to the study by Martin et al. [114] and choose the data of the entire employed population of China at the end of 2022 as the reference data in our study to test the representativeness of the sample. The reference data is sourced from the China Population and Employment Statistical Yearbook 2023 (https://www.zgtjnj.org/navibooklist-n3024011706-1.html). It is worth noting that, comparing the profile of the respondents with that of the entire working population in China profile, we find that the sample is slightly more educated, with a higher percentage of individuals with bachelor’s degrees compared to the entire employed population of China; furthermore, the age distribution of the sample is more youthful, with the majority of the group under 45 years of age, while the percentage of the group over 50 years of age is significantly lower than the percentage of the national employed population. This result suggests that our respondents are generally more educated and more skewed toward a younger age group. In light of the fact that millennial workforce has become the largest group in the labor market at the present time [115], and also given precedents set by related research [116–118], we believe that such a discrepancy is reasonable and consistent with the topic of our current study.

Measurements

Except for the measurement of the digital-HRM variable, which was measured using a self-developed scale, the scales selected for the measurement of the remaining variables in this paper are more mature scales that have demonstrated high reliability and validity in previous studies. We used a five-point Likert scale to measure all variables from “strongly disagree” (1) to “strongly agree” (5).

For the current study, digital-HRM was measured by 6-items scale that we developed ourselves, example items included “My organization uses digital technology to analyze and make decisions on assessment-related data.” Perceived ease of technology use was measured through Davis’s 4-item scale [101], example items included “Learning to operate the digital technology is easy for me.” Relative deprivation was measured using 4-items scale adapted from Cho et al. [119], example items included “When I compare what I have with my colleagues, I feel like I’m missing something.” Sense of work gain was measured using 8-items scale adapted from Zhu and Liu [65], example items included “Work has allowed me to become proficient in job knowledge and skills.” Cyberloafing was measured using 6-items scale adapted from Blau et al. [120], these items included “Receiving non-work related emails during working hours.” Employee innovative performance was measured using 6-items scale developed from Janssen and Van Yperen [121], these items included “I offer new ideas to improve existing working conditions.” Specific measurement items for each variable can be found in S1 Appendix.

Additionally, we included five control variables, namely age, gender, education, work experience and firm ownership.

Techniques for data analysis

The study uses the partial least squares structural equation modeling (PLS-SEM) technique. The SmartPLS software was employed. To test the moderated mediation effect (H4a and H4b), the PROCESS macro suggested by Hayes [122] was used. For the moderated mediation was tested by using Model 7.

Demographic profile of participants

The current study implemented the survey approach to collect data from respondents of various age brackets, gender, education, work experience, and firm ownership. The data collected from the respondents depicted 44.4% of the individuals as males and 55.6% as females. In terms of age bracket, 39.6% of the individuals were aged between 20–30 years, 54% were aged between 31–45 years, 5.8% were aged between 46–55 years, and 0.6% were aged 56 years and above. In terms of education level, 3.1% of the respondents are in high school or technical secondary school and below, 11.3% have a junior college degree, 69.2% have an undergraduate degree, and 16.4% have a graduate degree or above. In terms of work experience, 56.1% of all respondents have 4–10 years of work experience, while 27.5% have only three years or less, and 16.4% have more than ten years of work experience. In terms of firm ownership, 29.6% of the respondents were in state-owned enterprise, 62% were in private Chinese enterprise, 6.8% were in wholly foreign-owned enterprise, and 1.6% were in other ownerships, such as joint ventures (See Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographics characteristics of respondents.

https://doi.org/10.1371/journal.pone.0307195.t001

Descriptive statistics and correlations

Tables 2 and 3 shows the means and standard deviations of and the correlations among the study variables. All of control variables are related to our main variables respectively, so they were controlled for our main variables in the subsequent analysis. The results of the correlation analysis provide preliminary support for our hypotheses.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0307195.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Correlations.

https://doi.org/10.1371/journal.pone.0307195.t003

Assessment of common method bias

Before assessing the research model, we first assessed the common method bias using the full collinearity test. The result showed that the variance inflation factor (VIF) values of all constructs ranged between 1.150 and 2.582 (below 3.33; [123]). In addition, we adopted Harman’s single factor analysis to test common method variance [124]. The results show that the first principal component explained only 32.670% of the variance variation, which was less than the 40% threshold. Thus, we concluded that there is no serious homology variance problem in the current research.

Assessment of measurement model

It was verified that all reflective constructs presented a Cronbach’s alpha (α) and composite reliability (CR) greater than 0.7 [125]. Thus, construct reliability is adequate. Convergent validity was checked against average variance extracted (AVE). As reported in Table 4, the latent variables reach convergent validity because their AVE measurements exceed 0.5 [126]. Also, as shown in Fig 2, some factor loadings were below 0.7 but higher than 0.5, according to Hair Jr et al. [127], an item with outer loading >0.50 should not be removed if the construct achieved satisfactory values for AVE (>0.50) and internal consistency (>0.70), thus confirming reliability of the constructs.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Factor loadings.

https://doi.org/10.1371/journal.pone.0307195.g002

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Measurement model: Discriminant validity.

https://doi.org/10.1371/journal.pone.0307195.t004

Discriminant validity was checked with the Heterotrait-Monotrait (HTMT) ratio correlation [128]. Based on Table 4, all the HTMT values are lower than the conservative threshold value of 0.85, indicating discriminant validity is established between all the constructs. Overall, the model provides no reliability or validity concerns. In addition, model fit was assessed using root mean square residual (SRMR), a SRMR value of 0.054 showed an acceptable model fit.

Assessment of structural model

In the structural model evaluation (Table 5), the criteria to be considered include the assessment of possible collinearity problems through the determination coefficient (R²) and the blindfolding-based cross-validated redundancy measure (Q²). As can be seen in Table 4, R² for cyberloafing = 0.197, R² for employee innovative performance = 0.430, R² for relative deprivation = 0.292, R² for sense of work gain = 0.495. The values for the Stone-Geisser Q² statistic presented for the four endogenous constructs are above zero. In consequence, the structural model in this study was acceptable.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Structural model: Evaluation indicators.

https://doi.org/10.1371/journal.pone.0307195.t005

Hypothesis testing

Path analysis with a 5000-sample bootstrapping run was performed to test the proposed hypotheses. Analysis results are shown in Table 6 and Fig 3.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Results of partial least square SEM model.

https://doi.org/10.1371/journal.pone.0307195.g003

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Results of hypothesis testing.

https://doi.org/10.1371/journal.pone.0307195.t006

Firstly, digital-HRM was positively associated with sense of work gain (β = 0.582, p < 0.001), with the latter also positively influenced employee innovative performance (β = 0.595, p < 0.001). The indirect effect of digital-HRM on employee innovative performance through sense of work gain was also significant (β = 0.346, p < 0.001), providing support for H1. Similarly, digital-HRM was negatively associated with relative deprivation (β = -0.324, p < 0.001), with the latter positively influenced cyberloafing (β = 0.342, p < 0.001). The indirect effect of digital-HRM on cyberloafing through relative deprivation was also significant (β = -0.111, p < 0.001), providing support for H2.

Secondly, results showed that perceived ease of technology use moderated the relationship between digital-HRM and sense of work gain, indicated by regression coefficient the interaction term on sense of work gain (β = 0.123, p < 0.01, see Fig 3) in that the relationship is stronger when perceived ease of technology use is high (see Fig 4). At the same time, perceived ease of technology use also moderated the relationship between digital-HRM and relative deprivation, indicated by regression coefficient the interaction term on relative deprivation (β = -0.105, p < 0.01, see Fig 3) in that the relationship is stronger when perceived ease of technology use is high (see Fig 5). Hence, H3a and H3b are verified.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. The interactive effect of digital-HRM and perceived ease of technology use on sense of work gain.

https://doi.org/10.1371/journal.pone.0307195.g004

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. The interactive effect of digital-HRM and perceived ease of technology use on relative deprivation.

https://doi.org/10.1371/journal.pone.0307195.g005

Finally, in order to test the moderated mediation effects (H4a and H4b), we used PROCESS technique. Results showed that perceived ease of technology use moderated the relationship between digital-HRM and employee innovative performance via sense of work gain (conditional indirect effect (+1 SD) = 0.3678, bootSE = 0.0628, 95% confidence interval (CI) [0.2521,0.4983]; conditional indirect effect (M) = 0.3011, bootSE = 0.0477, 95% CI [0.2133,0.4012]; conditional indirect effect (- 1 SD) = 0.2344, bootSE = 0.0396, 95% CI [0.1637,0.3192]), the 95% CI in the difference of these effects excluded zero (difference = 0.1333, 95% CI [0.0466, 0.2197]). Overall, H4a is supported.

Perceived ease of technology use also moderated the relationship between digital-HRM and cyberloafing via relative deprivation (conditional indirect effect (+1 SD) = -0.0969, bootSE = 0.0245, 95% CI [-0.1518,-0.0545]; conditional indirect effect (M) = -0.0735, bootSE = 0.0176, 95% CI [-0.1126,-0.0438]; conditional indirect effect (- 1 SD) = -0.0502, bootSE = 0.0144, 95% CI [-0.0817,-0.0259]), the 95% CI in the difference of these effects excluded zero (difference = -0.0467, 95% CI [-0.0872, -0.0132]). Overall, H4b is supported.

Robustness test

To further test the hypothesis results, we conducted a series of robustness tests.

First, we reanalyzed the data after removing all control variables. Using SmartPLS for analysis, the results, as shown in Table 7, still support the proposed hypothesis. Also, using PROCESS to test the moderated mediation effect, the results indicate that when the perceived ease of technology use is high, the indirect effect of sense of work gain is 0.4044, with a 95% CI [0.2816, 0.5359]; when the perceived ease of technology use is low, the indirect effect of sense of work gain is 0.2534, with a 95% CI [0.1809, 0.3382]. The difference in indirect effects between high and low conditions is 0.1510, with a 95% CI [0.0609, 0.2457], excluding 0, supporting H4a. When the perceived ease of technology use is high, the indirect effect of relative deprivation is -0.1094, with a 95% CI [-0.1675, -0.0642]; when perceived ease of technology use is low, the indirect effect of relative deprivation is -0.0539, with a 95% CI [-0.0862, -0.0282]. The difference in indirect effects between the high and low conditions is -0.0555, with a 95% CI [-0.1005, -0.0181], excluding 0, supporting H4b.

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Results of hypothesis testing.

https://doi.org/10.1371/journal.pone.0307195.t007

Secondly, we performed a second test of the mediation and moderation hypotheses using hierarchical linear regression via SPSS software. As shown in Tables 8 and 9, the maximum variance inflation factor (VIF) coefficient of all models was 2.157, which was less than the threshold of 5 [123, 129], so no serious multicollinearity was found. Moreover, the F value of all models reached the level of significance (p <0.01), indicating that the model was statistically significant.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. Results of hierarchical linear regression analysis.

https://doi.org/10.1371/journal.pone.0307195.t008

Download:

• PNG
  larger image
• TIFF
  original image

Table 9. Results of hierarchical linear regression analysis.

https://doi.org/10.1371/journal.pone.0307195.t009

According to Baron and Kenny [130] and Liu and Zhang [131], we utilized the three-step regression method to test whether SWG and RD has a mediating effect. Step 1: DHRM (M5/M12) has a direct effect on EIP/CL. Step 2: DHRM (M2/M9) has a direct effect on SWG/RD. Step 3: After SWG/RD is introduced, the effect of DHRM on EIP/CL still exists. As shown in M7/M14, p < 0.001 with statistical significance. With DHRM, the β coefficient dropped/rised to 0.294/ -0.243, indicating that SWG/RD plays a positive/negative partial mediates effect relationship between DHRM and EIP. To sum up, H1 and H2 was supported.

From M3 in Table 8, when sense of work gain is used as the dependent variable, perceived ease of technology use and the interaction term between digital-HRM and perceived ease of technology use are added to the basis of M2. The results show that perceived ease of technology use positively moderates the relationship between digital-HRM and sense of work gain (β_DHRM = 0.575,p<0.001; β_DHRM×PETU = 0.182, p<0.001), thus confirming H3a. Similarly, from M14 in Table 9, when relative deprivation is the dependent variable, perceived ease of technology use and the interaction term between digital-HRM and perceived ease of technology use are added to the basis of M9. It is found that perceived ease of technology use positively moderates the relationship between digital-HRM and relative deprivation(β_DHRM = -0.243,p<0.001; β_DHRM×PETU = -0.143, p<0.01), thus confirming H3b.

Finally, we use the PROCESS macro model and the Bootstrap to retest the mediation effects of sense of work gain and relative deprivation and the moderation effect of perceived ease of technology use for robustness, setting the sample size to 5000. If the 95% confidence interval does not include 0, it indicates that the results are significant and the hypothesis is further confirmed. For the robustness test of the mediation effects, as shown in Table 10, the indirect effect of sense of work gain has a 95% CI [0.2306, 0.4269], not including 0, suggesting that sense of work gain significantly mediates between digital-HRM and employee innovative performance. The indirect effect of relative deprivation has a 95% CI [-0.1367, -0.0575], not including 0, indicating that relative deprivation significantly mediates between digital-HRM and cyberloafing, further supporting the original hypothesis. As for the robustness test of the moderation effect, the results indicate that perceived ease of technology use significantly moderates between digital-HRM and sense of work gain(β = 0.2521, p < 0.001); it also significantly moderates between digital-HRM and relative deprivation(β = -0.2438, p < 0.01), as shown in Table 10. Under conditions of high perceived ease of technology use, the positive/negative impact of digital-HRM on sense of work gain/relative deprivation is enhanced, thereby further supporting the original hypotheses.

Download:

• PNG
  larger image
• TIFF
  original image

Table 10. Results of mediation effect and moderated mediation effect.

https://doi.org/10.1371/journal.pone.0307195.t010

Although this study utilized a three-stage time-lagged method to collect data, it could not assert causality between variables. Therefore, building on the original measurement tools and data, we conducted a second follow-up survey of 487 respondents from November 24, 2023 to November 30, 2023, using the same questionnaire as the first survey. A total of 248 questionnaires were collected. We used IP addresses to match the questionnaires from the two collections, and after eliminating incomplete and uniform responses, a total of 241 valid questionnaires were obtained.

Before testing, we used the χ², CFI, TLI, and RMSEA to evaluate the model’s fit. Compared to the χ² value, the other fit indices are less influenced by the sample size and account for both the goodness of fit and the model’s complexity. For the model’s fit, χ² = 203.093, CFI = 0.946, TLI = 0.907, and RMSEA = 0.064, all of which are at satisfactory levels.

The structural coefficients estimated are presented in Table 11. According to the analysis, when controlling for stability effects (i.e., the influence of variables measured at Time 1 on the same variables measured at Time 2), Time 1’s Digital-HRM still significantly affects Time 2’s sense of work gain(β = 0.191, p<0.01) and Time 2’s relative deprivation(β = -0.126, p<0.05). This finding suggests that there is a causal relationship between digital-HRM and sense of work gain, as well as between digital-HRM and relative deprivation, supporting our predictions.

Download:

• PNG
  larger image
• TIFF
  original image

Table 11. Results of cross-lagged analysis.

https://doi.org/10.1371/journal.pone.0307195.t011

Likewise, after controlling for stability effects (i.e., the impact of sense of work gain at time 1 on sense of work gain at time 2, and the impact of employee innovative performance at time 1 on employee innovative performance at time 2), the impact of sense of work gain at time 1 on employee innovative performance at time 2 was found to be significant(β = 0.458,p<0.001). this indicates that sense of work gain is a cause leading to employee innovative performance, consistent with our predictions. however, after controlling for the stability effects (i.e., the impact of relative deprivation at time 1 on relative deprivation at time 2, and the impact of cyberloafing at time 1 on cyberloafing at time 2), the cross-lagged effect of relative deprivation at time 1 on cyberloafing at time 2 was not significant(β = 0.031,p>0.05), failing to support our prediction.

For the reason behind these findings, we speculate that it may be due to the first data collection occurring at the end of December 2022, during the peak of pandemic control measures. Most employees were in home isolation working online, without supervision from supervisors or colleagues, thus having more freedom and opportunity to engage in cyberloafing. However, the second data collection was conducted at the end of November 2023, when the pandemic control period had ended, and work had returned to normal order. Most employees were working in offices with supervision, significantly reducing the space and freedom for cyberloafing. Even if they felt deprived, they might not dare to engage in cyberloafing. Additionally, this period was marked by economic recovery, with many companies adjusting and getting back on track, requiring employees to perform well to avoid layoffs. Therefore, this could lead to the nonsignificant relationship between relative deprivation and cyberloafing.

Additionally, we have made a new discovery: Time 1’s sense of work gain significantly positively impacts the increase of digital-HRM at Time 2. this suggests that an increase in employee’s sense of work gain can further promote the practice level of digital-HRM in enterprises.

Theoretical implications

Our study makes three significant theoretical contributions. Firstly, our study responds to the academic call to identify and understand the newly emerging job resources in contemporary work environments [133]. The mainstream perspectives on digitalization and digital technologies often associate them with new forms of Tayloristic management, such as digital control, algorithmic supervision, and algorithm management [12]. In contrast, our research utilizes the job demands-resources theory as a framework, revealing the unique role and potential value of digital-HRM in actual work settings. We emphasize that digital-HRM is an emerging job resource adapted for the digital era, not an oppressive job demand. Specifically, digital-HRM meets employees’ needs for flexibility, skill development, and personalized services by providing a digitalized workplace, digital employee services, and digital HRM processes. This not only enriches the existing literature on the job demands-resources theory but also expands the scope of what constitutes a job resource, thereby aiding in enhancing employee engagement and performance.

Secondly, from a people-centered perspective, we explored the mechanisms by which digital-HRM affects employee-level factors (employee innovation performance and cyberloafing) based on the HRM-WB-OIP theoretical model. We verified that digital-HRM influences employee psychology in a mutually beneficial manner, thereby enhancing performance outcomes. This is the first empirical paper to validate how digital-HRM enhances performance by modulating employees’ psychological states, theoretically expanding the research on digital-HRM in an individual context and contributing to the existing literature on digital-HRM. In recent years, digital-HRM has garnered significant academic attention. Most existing studies discuss digital-HRM from the perspective of “things”, such as the conceptual framework of digital-HRM [53], its implementation paths [33], or the consequences of digital HRM at an organizational level, like organizational performance [37] and productivity of HR functions [36]. However, surprisingly, few studies focus on the impact of digital-HRM on individual-level factors, neglecting attention to the “people”, who are the main subjects of digital-HRM. While some literature has explored the impact of digital-HRM on employee performance [35], it is noteworthy that technology has social attributes [141]. If the implementation of digital-HRM focuses solely on employee performance without considering changes in their psychological states or exploring how to help employees better adapt to digital-HRM, it may not only damage the relationship between the organization and its employees [142] but also hinder the success of digital-HRM, ultimately affecting organizational development negatively [143]. Therefore, it is imperative to deeply understand how digital-HRM influences employee performance by affecting their psychological states. The introduction of the job demands-resources theory offers a new perspective. Relying on this theory, our study posits that digital-HRM is an emerging job resource, not a job demand. It empirically validates that digital-HRM impacts employees’ psychological states and behaviors in a mutually beneficial manner. Specifically, digital-HRM enhances positive psychological experiences (i.e., sense of work gain) and reduces negative psychological experiences (i.e., feelings of relative deprivation), boosting employees’ innovative performance and mitigating cyberloafing behaviors. This deepens our understanding of how digital-HRM serves as a critical factor influencing changes in employee psychology and behavior. This understanding is crucial for organizations, as it guides them on how to more effectively design and implement digital-HRM strategies to promote employee psychological well-being and the long-term health of the organization. Additionally, by introducing the concept of “sense of work gain [65]”—a derivative from a Chinese indigenous psychological framework—our study not only aligns more closely with the Chinese context but also significantly expands the theoretical horizon of digital-HRM, revealing how it enhances employee well-being and drives performance improvements. This also enriches the global discussion on the intersection of technology and employee psychology.

Thirdly, our study contributes to understanding the boundary conditions that affect the effectiveness of digital technology-related job resources, particularly those related to employees’ individual capabilities and beliefs. Our findings indicate that employees’ perceived ease of technology use are a critical factor influencing how they utilize and respond to digital-HRM. Employees with a higher perception of technology ease of use are more likely to actively accept and utilize these resources, as they believe that technology simplifies workflows, enhances work efficiency, and reduces stress caused by technological complexity [144]. Consequently, these employees exhibit a greater sense of work gain and a lesser feeling of relative deprivation when interacting with digital-HRM practices. By identifying this boundary condition, our study enhances the exploration of factors affecting the effectiveness of digital technology-related job resources and offers a new theoretical perspective on the relationship between digital-HRM and employees’ psychological responses. Our results show that when employees have a positive perception of technology ease of use, the relationship between digital-HRM and their sense of work gain and relative deprivation becomes more pronounced, providing crucial insights for organizations in designing and implementing digital-HRM strategies.

Practical implications

Our research also has significant practical implications for organizations. First, we suggest that organizations consider the benefits of digital transformation and carefully evaluate the implementation of digital-HRM initiatives based on their specific organizational context and readiness. The field of HRM is undergoing rapid changes, and it is increasingly important for HRM not just to serve as a support function, but to actively participate in shaping organizational strategies through digitalization [40]. Organizations are therefore encouraged to develop digital and intelligent HRM systems tailored to their unique needs, which can maximize their potential to enhance employee performance and offer employees greater opportunities for growth. This thoughtful and strategic approach will enable organizations to effectively navigate the evolving competitive landscape and maintain their long-term competitiveness.

Second, from an HR perspective, the increasing adoption of digital technology enables HR professionals to execute various HRM functions more efficiently and swiftly. This encompasses activities such as recruitment processes, training programs, job evaluation, performance measurement, compensation management, and employment relations [39]. To effectively facilitate HRM processes and digitalization, future HR managers will need to acquire digital knowledge, skills, and competencies. This will enable them to proficiently identify and match talent, utilize digital tools like talent portraits and talent dashboards, and provide technical support for talent acquisition, growth, and development within the organization, Such as IBM’s application Watson Recruitment, which recommends the best candidate for any job based on job requirements, Blue Match software, which drives employee career development by algorithmically providing each employee with suggestions for career advancement and new jobs and MYCA, which helps employees identify directions for skill enhancement [3], in addition to Deloitte’s talent management system [145], and AsiaInfo’s Talent Training Programme [146]. Furthermore, to ensure the success of digital transformation initiatives, it is crucial to fully meet the needs of employees [147]. When integrating new digital and artificial intelligence technologies, the human resources department should design and implement a range of targeted technology training programs, such as skill enhancement, online interactive courses, and certification training, to accommodate the learning preferences and career development needs of various employees. Regular, interactive training sessions can help employees demystify these technologies, reducing uncertainty and resistance to new systems. This approach not only helps employees perceive digital tools as enablers rather than threats and builds trust with managers [1] but also stimulates a positive attitude towards technological changes, enabling them to fully realize their potential in new work models. Additionally, organizations can foster a digital-centric corporate culture through cultural activities and innovative training programs, which will aid in developing digital mindsets, innovation awareness, and team collaboration spirit among employees. This encourages them to proactively adapt to changes, actively explore new working methods, and promote ongoing innovation and improvement within the organization.

Last, with regards to individual employees, the digitization of HRM is poised to transform employee-related functions, such as recruitment, training, and assessment, necessitating a profound comprehension of the ensuing changes [40]. It is incumbent upon employees to proactively cultivate relevant digital technology competencies, nurture a sense of self-efficacy in utilizing digital technology, foster a mindset of empowerment rather than being subject to digital control, and be prepared for long-term collaboration with digital technology and AI. Rather than evading or shunning these advancements, employees should embrace this challenge and equip themselves accordingly.